Home › Categories
Browse AI Prompts by Category
Every prompt here is free, tested, and written with fill-in variables so you can make it yours in about ten seconds. Pick the category that matches the job you are doing rather than your job title. That single habit tends to produce better output than any amount of prompt tweaking.
PromptCraft Asia organises 556 free AI prompts across 20 categories. The core ten are business, writing, coding, career, social media, education, email, image generation, personal productivity, and SEA business. Ten focused collections cover finance, legal, marketing, data analysis, HR, customer service, entrepreneurs, teachers, Canva AI, and Notion AI. No signup needed.
Core Categories
The main library, covering the work most people come here for.
Business
Strategy, planning, proposals, and the everyday writing that keeps a business moving.
Writing
Drafting, editing, and rewriting, from blog posts to the email you have rewritten four times.
Coding
Debugging, code review, refactoring, and explaining what someone else's code actually does.
Career
Resumes, cover letters, interview prep, and the awkward conversations about pay.
Social Media
Captions, hooks, content calendars, and turning one idea into a week of posts.
Education
Study aids, explanations, revision plans, and breaking hard topics into something learnable.
Outreach, follow-ups, replies, and saying no politely without writing six drafts.
Image Generation
Prompts for Midjourney, DALL-E, and the rest, with the structure that gets usable results.
Personal Productivity
Planning, prioritising, note-taking, and getting your week back under control.
SEA Business
Written for Southeast Asia, with local platforms, payment habits, and multilingual customers assumed.
Focused Collections
Narrower sets for specific roles and tools.
Marketing
Campaigns, positioning, ad copy, and messaging that says something specific.
Finance and Accounting
Budgets, forecasts, reporting, and explaining numbers to people who dislike numbers.
Legal and Compliance
Contract review, policy drafting, and plain-English summaries of dense documents.
HR and Recruitment
Job descriptions, screening, onboarding, and feedback that is honest without being brutal.
Customer Service
Replies, escalations, refunds, and handling a complaint without sounding like a robot.
Data Analysis
Cleaning, summarising, spotting patterns, and turning a spreadsheet into a finding.
Entrepreneurs
Idea validation, pitching, early hiring, and the decisions nobody warns you about.
Teachers
Lesson planning, marking, differentiation, and reclaiming your Sunday evening.
Canva AI
Design and content prompts written specifically for Canva's AI features.
Notion AI
Templates and workflows for getting real work out of Notion AI.
New to prompting? Start with how to write better AI prompts, then look at few-shot prompting and role prompting. Those three cover most of what separates a prompt that works from one that nearly works.
How the Categories Are Organised and How to Choose
Categories are grouped by the kind of work rather than by industry, because the same prompt structure serves a marketer writing ad copy and a founder writing a landing page. Broadly they fall into four families: communication categories covering writing, email and social media; commercial categories covering marketing, business, finance and customer service; specialist categories covering coding, data analysis, legal and HR; and tool specific categories for Notion AI, Canva AI and image generation.
Choose by the output you need rather than the label that describes you. If you want a document, look at writing. If you want something persuasive aimed at a customer, look at marketing. Where two categories seem plausible, pick the narrower one, since specific prompts outperform general ones.
The SEA business category exists because regional context changes the output rather than only the tone, covering multilingual audiences, local payment norms and professional conventions. The OECD AI Policy Observatory tracks national policy on AI adoption across countries, which is useful background if you are deploying prompts in a regulated setting.
Why These Prompt Categories Work
A category is useful when it groups prompts by the shape of the output rather than by the person requesting it. A founder, a marketer, and a freelancer writing the same landing page need the same prompt structure, so sorting by job title would scatter one technique across three places. The categories here are organized around what comes back: a document, a persuasive message, a structured dataset, code, or an image. That makes the right starting point easy to find even when your role fits no single label.
Regional context sits inside the categories rather than replacing them. A marketing prompt used in Jakarta and one used in Sydney share their structure and differ in the variables you fill in, which is why the fix is usually naming the market, currency, language, and platform rather than finding a different prompt. The SEA Business category exists for cases where regional detail is the substance of the task, such as marketplace listings or local tax explanations. For everything else, add two lines of local context.
The most common mistake when browsing is picking the broadest category. Specific prompts outperform general ones consistently, so choose the narrower of two plausible options. The second is treating a category as a menu to read rather than a set of templates to run, since you learn more from adapting one prompt than from skimming twenty. The third is ignoring the fill in variables and sending the prompt as written, which produces exactly the generic output those variables exist to prevent.
How to Use These Prompts
- Start from the output you need. Decide whether you want a document, a short persuasive message, structured data, code, or an image, then open the category that produces that thing.
- Choose the narrower category when two fit. A prompt written for email outreach beats a general writing prompt for the same task, because its defaults are already set for that format.
- Replace every bracketed variable before running the prompt. The variables are where your context enters, and leaving them in place is the single biggest cause of generic output. Fill them in every time.
- Add regional detail as a separate line. Market, currency, language, and platform take one sentence to specify and change the result more than any rewording of the main instruction. It is the cheapest available improvement.
- Save the prompts that work for you. A personal set of ten tuned prompts is worth more than the full library, because you already know how each one behaves in practice.
What Results to Expect
Expect a strong first draft from most categories and a finished artifact from almost none. Prompts reduce the cost of starting and impose a sensible structure, while judgment, accuracy, and voice remain yours. Results are best on tasks with a clear format and a well defined audience, and weakest on anything requiring current facts or internal knowledge the model has never seen. Anthropic guidance on prompt engineering frames the same tradeoff, favoring explicit instruction over clever phrasing. Verify anything time sensitive yourself before you publish it.
Iterate by changing one variable at a time so you can attribute the difference to something. If an output misses, describe the gap concretely rather than asking for an improved version. Different task types respond to different techniques, a point catalogued across dozens of methods in a systematic survey of prompt engineering, so a method that transforms your code prompts may do nothing for your marketing ones. Test within a category before assuming it generalizes, and assume nothing transfers until you have checked.